Yahoo Finance

Aaron Levie

Thanks, Cynthia, and thank you all for joining the call today.
We delivered strong Q4 operating results, reflecting stabilization of IT budgets and continued growth in customer demand for Box AI. We achieved revenue of $280 million, up 6% year over year, or 8% in constant currency, operating margins of 27.3% and EPS of $0.42, $0.01 above our guidance.
In fiscal 2025, we drove revenue growth of 5% year over year, or 7% in constant currency and expanded operating margins by 320 basis points to 28%. It was a defining year for Box as we introduced the most transformational product lineup in our company's history with the delivery of our intelligent content management platform.
In the last two weeks of the quarter, we released Enterprise Advanced, combining the full power of our ICM platform to customers in a single multiproduct offering. We are pleased with the momentum we are seeing with customers and close several dozen Enterprise Advanced deals in Q4. We are seeing companies start to adopt Enterprise Advanced to power intelligent metadata extraction from documents, automate workflows and dashboards with Box Apps, gain access to Forms, Dock Gen and Archive, and create custom AI agents with the AI studio.
Examples of Enterprise Advanced customer wins in Q4 include, through a systems integrator, a state DMV who is forecasting a huge increase in online title submissions due to changing regulations will be using Enterprise Advanced to replace a legacy file system and to create custom portals to ease customer submissions and extract metadata to create dashboards for processing and managing of records.
A US-based global investment firm that uses Box to power its client portal as well as provide a custom internal document management system for its financial advisers upgraded to Enterprise Advanced to support contract life cycle management for their legal team, Doc Gen for multiple departments, and AI Studio to deliver content-focused AI agents to help advisers answer questions faster about their documents and enhance their workflows.
Also through a systems integrator, an American international law firm moved to Enterprise Advanced to expand their use of Box with their organization. Specifically, their information governance team is looking at ways to leverage Box AI to help automate internal processes as they deal with repeated documents from templates. They will also continue to build out high-value use cases over the next several months from the Box AI studio using metadata extraction on contracts, archive and more.
As we've seen in our Q4 deals and in talking with our customers throughout the past year, we're clearly entering one of the biggest shifts in business that we've ever seen, driven by AI. AI agents are entering the workforce and will augment and accelerate our work.
Our own structured data is enabling intelligence that we can now use to gain new insights about our business and we can begin automating any workflow in the enterprise, especially that long tail of work that we couldn't automate before. The companies that take advantage of this moment will be the ones that win in the future. And of course, the ones that don't will fall behind and be severely at risk.
At Box, internally, we are building an AI-first enterprise. We are looking at every aspect of our business and imagining what we can do to turbocharge our productivity from our employees, allowing us to move faster, making more informed decisions, automate more busy work and overall, just get more done.
Box AI is central to this effort. As we use Box AI and hubs to help Box employees ask HR sales enablement questions to get information faster. Box engineers -- we use Box AI in notes to review code for bugs or write documentation. In go-to market, we use Box AI to write sales pitches and proposals for customers. And in fact, I even use Box AI to help improve my remarks today.
Additionally, beyond Box AI, we're leveraging technology from key partners to help in engineering, customer success and more. And most importantly, we are bringing this exact same value to all of our customers so they can begin to operate in an AI first way as well.
For the first time ever, organizations can truly tap into the value of their enterprise content. All of a sudden, any employee can gain the same level of expertise as the most knowledgeable employee just by asking questions of the existing enterprise data that's already there. The hidden information inside of contracts, invoices, financial documents and customer data turns into business critical insights that drive better execution.
Agentic AI can automate workflows that were expensive and time-consuming and by understanding within our content, we can better secure and protect it at scale. This is the power of intelligent content management from Box to help our customers transform how they work with content and AI.
As noted, we officially launched our new Enterprise Advanced plan at the tail end of Q4. With Enterprise Advanced, our customers can build no-code apps with Box Apps, leverage Box Forms and Dock Gen to automate workflows, use Box AI metadata extraction to plot structured data from documents, create Box AI agents and with the Box AI studio be able to extend AI into more workflows and more.
These new capabilities represent the foundation of intelligent workflow automation within Box, enabling the automation of business processes like contract management, digital asset management, invoice processing, clinical trial management, loan origination and thousands of other workflows. And while FY25 was an important and critical year for us in delivering our intelligent content management platform, we are just getting started.
In FY26, we will be advancing our platform to transform how enterprises work with their enterprise content and AI. We have many incredible product announcements in store to help our customers extract critical structured data from documents with AI, leveraging our Alphamoon acquisition from last year.
We'll continue to double down on Box Apps to power no-code applications, deliver major advancements to Box Relay for workflow automation and more. We'll also continue to strengthen our security and compliance posture with new breakthroughs in Box Shield, the launch of Box Archive and more to help customers protect and govern their sensitive content.
Finally, we'll be doubling down on our core AI platform capabilities, including more powerful intelligent document processing, agentic workflows, exploring new forms of data classification, better support for Box AI agents and more. And as we've seen in recent weeks, the rate of innovation from AI model providers like OpenAI, DeepSeek, XAI, Gemini from Google and Anthropic is only accelerating.
Just in the past two weeks, we've made new models like GPT 4.5 and Claude 3.7 Sonnet available to customers with access to the Box AI studio. Both of these new models have shown dramatic improvements on reasoning, math and coding skills, which is extremely valuable in processing complex documents, contracts, lease agreements, financial records and more. In fact, in our Box AI Enterprise eval, we found that GPT 4.5 performs a whopping 20 points better than GPT 4.0 did for extracting data from documents in a single shot. These are huge advancements that will help drive more accuracy and quality to help enable even more mission-critical workflows with AI.
To enable customers to experience Box AI more easily as they grow with Box, in February, we announced that we expanded Box AI availability in our business, Business Plus and Enterprise plans. And in addition to offering higher tiers of AI value within Enterprise Plus and Enterprise Advanced, we know that many customers have high-volume use cases for using AI on their content. And to help enable these use cases, we've also added an AI consumption-oriented model to our platform with the launch of Box AI units.
These AI units allow customers to leverage AI credits for high-volume tasks like metadata extraction from documents, leveraging more Advanced AI models as they get released, creating custom AI agents to automate workflows in the future and more.
And finally, we have been thrilled with the recent recognition of our industry-leading intelligent content management platform by several independent analyst firms. This includes being named a leader in the Forrester Wave for content platforms in Q1, a leader in the Gartner Magic Quadrant for Document Management last year and a leader in the IDC MarketScape for worldwide Intelligent Content Services in 2024.
Now turning to go-to market. We continue to enable new and existing customers to recognize the full value of Box's ICM platform with increased adoption of our multiproduct offerings. Suites represented 87% of deals over $100,000, up from 81% a year ago. We saw continued solid suites attach rates in large deals across all geographies. We grew suites to 60% of total revenue in Q4 compared to 55% a year ago.
In addition to the exciting announcements that we're seeing around our newest suite, Enterprise Advanced, we still have a large opportunity to drive Enterprise Plus adoption, driven by our enhanced Box AI solutions. As we bring intelligent content workflows to the enterprise with AI, going to market with partners is also going to become increasingly critical.
We continue to expand our work with key partners and system integrators across the world that can bring Box to more enterprises and embed us more deeply into our customers' workflows. We saw notable partner-led customer wins with Enterprise Advanced in Q4 as we drove larger deals with customers and powered more workflows in partnership with key system integrators.
Fiscal 2025 was a pivotal year for Box, culminating in the launch of Enterprise Advanced, delivering to customers the full power of our intelligent content management platform in a single offering. As I stated earlier, we are clearly entering one of the biggest shifts in business that we've ever seen, driven by AI. Our strong financial model provides us with the opportunity to execute on our robust product roadmap and invest in strategic go-to-market initiatives, leading to accelerating our revenue growth in the long term.
We will be providing more detail on our roadmap and growth strategy during our upcoming Financial Analyst Day on March 18.
And with that, I'll hand it over to Dylan.

Dylan Smith

Thanks, Aaron. Good afternoon, everyone, and thank you for joining us today.
In fiscal 2025, we made substantial progress against the three financial priorities we outlined heading into the year. First, we strengthened our foundation to deliver profitable growth by launching Enterprise Advanced and ramping up key go-to-market initiatives. Second, we significantly expanded gross and operating margin through cost discipline and advancing our workforce location strategy. Finally, executing our capital allocation strategy enabled us to further strengthen our balance sheet and efficiently return capital to shareholders.
In FY25, we delivered revenue of $1.09 billion, up 5% year over year and up 7% in constant currency. Operating margin came at 27.9%, up 320 basis points year over year and up 460 basis points in constant currency. We delivered EPS of $1.71, up 17% year over year and up 27% in constant currency. Finally, in FY25, we generated record free cash flow of $305 million, up 13% year-over-year.
Turning to Q4, we closed the year with strong results, exceeding our guidance for Q4 revenue and EPS. Q4 revenue of $280 million was up 6% year over year and up 8% in constant currency, exceeding our expectations despite absorbing an incremental FX headwind of 70 basis points. We now have approximately 1,920 total customers paying us at least $100,000 annually, up 8% year over year.
Our Q4 Suites attach rate in large deals was strong at 87%, an improvement from 81% in Q4 of last year. When we launched Suite several years ago, our objective was for suites to represent the vast majority of our sales to large customers, and we are pleased to have achieved this goal. As such, going forward, we will no longer be providing this metric.
As our intelligent content management platform enables customers to adopt Box for more Advanced use cases, our percentage of total revenue attributable to Suites will continue to grow. In Q4, 60% of our revenue was attributable to Suites customers, up from 55% a year ago. We ended Q4 with remaining performance obligations, or RPO, of $1.5 billion, a 12% year-over-year increase and up 14% in constant currency.
Our strong RPO growth continues to be driven by longer customer contract durations, demonstrating our customers' long-term commitment to Box and resulting in the long-term portion of Q4 RPO growing by 21% year-over-year.
Consistent with prior quarters, we expect to recognize roughly 60% of our RPO over the next 12 months. Q4 billings of $399 million were up 5% year over year and up 7% in constant currency. Q4 billings exceeded our expectations of a low single-digit growth rate due to solid bookings and a high volume of early renewals despite a minor FX headwind versus our prior expectations. We ended Q4 with a net retention rate of 102% in line with our expectations and up from 101% a year ago. Our annualized full churn rate remains at a best-in-class 3%, reflecting the stickiness of our high-value solutions.
Looking ahead, we expect our net retention rate to remain at 102% in Q1 and to improve to 103% by the end of FY26. Q4 gross margin of 81.0% was up 260 basis points year over year. Achieving this top-tier gross margin profile is the result both of a multi-year effort across our teams and of how product innovation has consistently increased our platform's differentiation over the past several years. Having completed the sale of our data center assets in Q3, this result represents our normalized annual gross margin expectations as we head into FY26. Q4 gross profit of $226 million was up 10% year over year, exceeding our revenue growth rate by 350 basis points.
In Q4, we continued to drive cost discipline across the business, which includes advancing our lower cost location strategy. We ended FY25 with approximately 470 full-time employees in Poland, a significant increase from 300 a year ago.
Going forward, we will continue to scale our footprint in Poland, both as an engineering center of excellence and is a critical hub for our G&A functions. We achieved record Q4 operating income of $76 million, up 9% year over year. Q4 operating margin of 27.3% improved by 60 basis points year over year despite absorbing an FX headwind of roughly 110 basis points.
As a result, we delivered EPS of $0.42 in Q4, which includes a negative impact from FX of approximately $0.03. I'll now turn to our cash flow and balance sheet. In Q4, we generated free cash flow of $91 million, up 12% from Q4 of last year. We also generated a 14% year-over-year increase in cash flow from operations, which came in at $102 million. In fiscal 2025, we continued to strengthen our balance sheet, and we ended Q4 with $724 million in cash, cash equivalents, restricted cash and short-term investments.
In Q4, we repurchased 1.3 million shares for approximately $43 million. For the full year of FY25, we repurchased approximately 7.6 million shares for approximately $212 million which represents roughly 70% of our free cash flow generation.
As of January 31, 2025, we had approximately $52 million of remaining buyback capacity under our current share repurchase plan, and our Board of Directors just authorized an additional $150 million increase to our share repurchase program.
With that, let me now turn to our Q1 and FY26 guidance. As a reminder, approximately a third of our revenue is generated outside of the US, with roughly 65% of our international revenue coming from Japan.
Before providing guidance, I would like to provide some context on our tax expenses going forward. As a result of releasing our US tax valuation allowance in Q4, we expect to recognize incremental noncash deferred tax expenses in the US, which will result in a non-GAAP EPS headwind throughout FY26. Our non-GAAP adjustments will also include the effect of income tax being calculated on our higher non-GAAP earnings.
Beginning in FY26, we are establishing a long-term non-GAAP tax rate of 27% which will be applied to our GAAP to non-GAAP reconciling items and reflected in our non-GAAP EPS guidance for Q1 and FY26. Please note that due to our significant NOL carryforwards and other tax credits, we expect to pay more modest cash taxes in FY26 with an estimated cash tax in the range of $12 million to $15 million.
For the first quarter of fiscal 2026, we expect Q1 revenue to be in the range of $274 million to $275 million, representing approximately 4% year-over-year growth at the high end of the range and 5% growth in constant currency. Our Q1 revenue guidance also includes a 120-basis-point headwind due to the leap year in FY25.
We anticipate our Q1 billings growth rate to be in the low to mid-teens range. This includes an expected tailwind from FX of approximately 300 basis points. We expect Q1 gross margin to be approximately 80%. Note that in Q1 of last year, our gross margin benefited by approximately 100 basis points due to data center equipment sales in the quarter. We expect our Q1 non-GAAP operating margin to be approximately 25% versus 26.6% a year ago.
There are three factors outside of our underlying business, which combined to create an expected 220 basis point headwind when comparing Q1 to the prior year. This includes 100 basis points from data center equipment sales, 80 basis points due to the leap year and an expected approximately 40 basis points due to FX. When adjusting for these factors, our Q1 non-GAAP operating margin guidance represents a slight year-over-year improvement.
We expect our Q1 non-GAAP EPS to be in the range of $0.25 to $0.26 as compared to $0.39 a year ago. This includes an expected headwind of approximately $0.01 from FX and $0.11 from incremental noncash deferred tax expenses. Weighted average diluted shares are expected to be approximately $151 million.
For the full fiscal year ending January 31, 2026. We expect revenue to be in the range of $1.155 billion to $1.16 billion, representing approximately 6% year-over-year growth. We expect a neutral impact from exchange rates based on present currency rates. We expect our FY26 billings growth to be approximately 7%, including a tailwind of approximately 30 basis points from FX. Please note that while we anticipate fluctuations in our billings growth rate between Q1 and Q2, we expect our overall billings growth rate in the first half of the year to be roughly in line with our annual billings growth rate.
We expect FY26 gross margin to be approximately 81%. When adjusting for the impact from data center equipment sales from Q1 to Q3 in FY25, this represents a year-over-year improvement of 40 basis points. We expect our FY26 non-GAAP operating margin to be approximately 28%. Year-over-year comparisons include a 60-basis-point negative impact from last year's data center equipment sales.
As Aaron mentioned, we have a significant opportunity to transform how enterprises work with their content. To take advantage of this opportunity, in the year ahead, we will be making methodical investments in key go-to-market initiatives to ensure we can reach customers at this critical technology juncture, which includes building out a strong partner ecosystem to expand our reach. We will also be enhancing our critical AI and workflow capabilities to deliver the leading intelligent content management platform to enterprises.
In FY26, we will continue to drive efficiency across the business through cost discipline, AI for workflow automation and our workforce location strategy. We expect to realize significant returns from the investments we're making this year, and we remain committed to delivering significant margin expansion over time.
We expect FY26 non-GAAP EPS to be in the range of $1.13 to $1.17 as compared to $1.71 in the prior year. This includes a $0.52 headwind from incremental noncash deferred tax expenses. Weighted average diluted shares are expected to be approximately $153 million.
We are proud of the solid results we delivered throughout FY25. Our ability to generate business bottle leverage while making strategic investments in our intelligent content management platform is paying off and we couldn't be more excited about the opportunities ahead of us with Enterprise Advanced and AI.
We look forward to providing more details at our Financial Analyst Day later this month. With that, Aaron and I will be happy to take your questions. Operator?

Question and Answer Session

Operator

(Operator Instructions) Brian Peterson, Raymond James.

Brian Peterson

With the billings number was great this quarter, I'd just love to understand the early momentum you've seen with Advance, was it better than you expected? And how do we think about the pipeline for that product as we enter fiscal year '26?

Aaron Levie

Yeah. So we were very happy with the results in Q4 on Enterprise Advanced. This is a plan that our customers didn't really even have much early heads up on. We were moving pretty quickly throughout the year. and announced it officially at BoxWorks.
And so that only gave customers really kind of 60 days, maybe at best to really dive in, understand the offering, understand the plan. And so to have dozens of deals at the end of Q4 that were Enterprise Advanced, we were very happy about. The momentum is certainly building in all of our customer conversations.
And as folks who have paid attention to our rollout of these plans know, there's definitely a transition period from one plan to the next. But very quickly, most of the sales conversations we're going to have will certainly start with the functionality that's in Enterprise Advanced and then obviously, customers can down select to Enterprise Plus or another plan if they don't need that automation or intelligent document processing, but we will start to veer more toward highlighting all those capabilities and those use cases to customers, you're going to see this quite a bit in all of our major announcements they'll tend to have some Enterprise Advanced capability or component in them throughout the year, and we're just going to drive more and more momentum.
So super happy with the results in Q4 and Enterprise Advanced, and we expect that to be a core part of our sailing motion. But still wanting to note that it's a ramp process, as you can imagine.

Brian Peterson

Aaron, you mentioned partners a couple of different times, As we are thinking about the mix of kind of direct versus indirect or however you want to say it, how should we think about that evolving going forward?

Aaron Levie

Yeah. So we have long been in -- mostly direct business other than in Japan, which is a very channel-heavy business, even though we still have direct reps involved in those deals. And there's a lot of lessons in Japan that we'd like to be able to bring to other regions.
And just naturally with the kind of use cases we're now going after where we are really going into the core of a customer's transformational workflows moving from legacy systems to the cloud, being able to start to automate more business processes, these are the types of workflows that need change management, there needs to be implementation. It usually is a broader part of a bigger strategy and there's now obviously very, very relevant partners in all of these these categories that help customers every single day and every single industry with those types of transformation.
So we want to be embedded in each of those partners. When they think about content, we want them thinking about Box and so we'll bring them into more deals as relevant, and we want to build flywheels with a lot of these top GSIs as well as boutique and even kind of regional firms that have a specialty in a particular domain.
At the Analyst Day, we'll certainly talk a little bit more about the overall partner and system integrator strategy. But we do expect that this is a more critical part of our go-to-market engine, where last year, we started to rev up the engine and now this year, I think it's going to be a -- you'll hear more from us on this part.

Operator

Michael Funk, Bank of America.

Michael Funk

First one, higher-level macro question, growing uncertainty with the macro environment last month or so, do you get any sense from your clients about uncertainty on their end for purchasing due to macro or shift in behavior?

Aaron Levie

Yeah. Certainly, very top of line question. It's an increasingly dynamic environment, as I think we've all seen every day, there's some type of change or reversion that keeps everyone on their toes. We're staying very close to our customers on the environment, and we want to continue to partner with every customer and every industry to help them drive more efficiency to help them drive more automation to help them get more intelligence from their business. So we believe we're an asset in this environment in dynamic times.
You want to have more leverage from your technology. You want to be able to retire more legacy systems, you want to automate more workflows. So we think we're in a very strong position there. But certainly, dynamic macro environment, and we're continuing to watch it very carefully.

Michael Funk

And on NRR, so maybe help me think of components to NRR as far as pricing contribution products and then change in seed count as we think about maybe potential for improvement moving forward.

Dylan Smith

Yeah. So on NRR, I would say, for the, call it, medium term and what's baked into our confidence in being able to improve that net retention rate from 102% to 103% year over year largely driven by the same types of dynamics that we've experienced over the past year. which is, to say, continued improvements in price per seat.
So pricing driving a lot of the customer expansion, particularly now that we have Enterprise Advanced as part of our portfolio. And then expect to see still pretty muted and minimal contribution from seat growth and then at the same time and kind of completing the kind of three factors on the overall kind of full churn rate, we also expect that to remain very strong and stable at 3%.
And then outside of NRR, you're not expecting to see a big shift in the contribution of expansion from our existing customers, which is represented in that NRR and the contribution to growth from kind of new customers buying Box for the first time. So longer term, I think that of those various factors, the one that we will hopefully -- and we expect to contribute more to growth is on the seat side, but we'll say that we do expect even steady state and longer term for pricing improvements to have be a bigger driver of net retention improvements rather than seat growth.

Operator

Steve Enders, Citi.

Steven Enders

I guess maybe just to start on the Advanced customer adoption that you've seen so far. I guess maybe what are kind of the core use cases that you're seeing those customers utilize for? How much of an uplift has the Advanced plan driven from the ACV growth on a like-for-like basis for those customers? Just any further details around that would be helpful.

Aaron Levie

Yeah. So I'll go through some of the use cases, maybe Dylan can provide some of the pricing color. There's going to be a few different kind of killer apps within Enterprise Advanced. It's the first time that we've released a new suite that had on day one the kind of breadth of functionality that this does. The Enterprise Plus offering was really built on the success of Shield and then we added more to it over time.
Enterprise Advanced has a little bit of something for the different kinds of of workflow and content management use cases that are out there that we've been missing. So Forms and Doc Gen, if you're doing a lot of of sort of form-based automation where data comes in, you need to be able to automatically generate a contract or an invoice or some other statement with that data. So we've seen a number of customers be very excited about that functionality.
We have unlocked with Archive, a lot of customers' use cases where they have large data sets that they want to be able to manage within Box. But we haven't historically given them the tooling to manage that content outside of really the end user experience. It's how Archive lets you go put a content that you want to be able to govern over a long period of time for a record management use case and that's now built in. So those are a few of the capabilities that have been exciting.
The really, really big ones are likely going to really revolve around this metadata extraction from documents and so that's our AI-powered document processing capabilities. So you take a contract or an invoice or a bank statement and inside of that document, today with inbox. Unless you can like tell us what's inside of it, we can't let you really, really automate much around that document unless you're kind of manually moving it through a workflow or again, giving us the metadata details about that file.
What we now can do with AI is we can extract that data for you. So you put the contract into Box and you want to pull out the renewal date, the party names, the different key clauses. Any other parameters and we can pull out that data, we put it into our structured database and then you can automate the workflow around that. So you can imagine the endless use cases around invoices, statements, contracts, resumes, anything that right now you want to be able to get intelligence from or automated workflow around.
We now have increasingly the capabilities, starting with our APIs and then the Alphamoon technology lets us build that out. at the end user side later this year. So that's going to be probably one of the bigger growth drivers.
And then the second one is our no-code app functionality. That's based on the Crooze acquisition now about 1.5 years ago. And so you can kind of see these things as interrelated. Once I have the data about my documents or my contracts or my invoices, now I want to have a dashboard to look at all the insights from that data or I want to be able to have a dashboard to understand where is a particular document in its process within the enterprise.
And so I want to be able to build a no-code application for that workflow. So we're seeing a lot of demand, especially for those two core capabilities, and we think that's just going to continue to accelerate.

Dylan Smith

Yes. And then to hit on some of the economics, you asked for the ACVs. We are pleased that the Enterprise Advanced deals that we've sold so far have achieved the target pricing uplift of 20% to 40% versus E Plus that we had outlined when announcing Enterprise Advanced and as you'd imagine, that's where virtually all of the deals we've sold so far are coming from current or previously, I guess, Enterprise Plus customers.
Still a bit too early to tell and speak to the deal trends either on overall contract values or on pricing comparisons for customers who are coming into enterprise Advanced as their first sale, but we will certainly kind of share what we're seeing there once we have more of those data points as well.

Steven Enders

Okay, great. No, that's helpful context there. And then on the go-to-market investments that you're making, can you just help us, I guess, think through, I guess, what's kind of needed to be built out still or whether is, I guess, the most kind of like direct investment coming on the go-to-market front?
And I guess, secondarily, just as we think about the unit economics of a customer coming in from a partner versus direct, how do those compare as you kind of lead into that partner-driven model?

Aaron Levie

Yeah. So on the go-to-market side, and I do want to note that we're being very methodical about the investments. It's a very surgical approach to what we're going to be adding resources to. But if you look at where we are taking the platform and the kind of customer conversations we're having, there's a few components that we believe will accelerate those conversations and be able to respond to more of the demand that's out there. So you're going to see incrementally more on the industry's efforts and kind of key industries that we're going after.
You can think financial services, life sciences, et cetera. We are doing more on the system integrator front as per the kind of last question. So you can imagine that we want to have the right kind of talent enabling the key system integrated partners around our platform and then bring them into more deals and really building that flywheel. So there's kind of an expansion of the kind of relationships we want to have there.
We are going to be investing in key segments of the business that simply need more sales capacity or the sort of respective overlay functions that are tied to that. And so those will be segments that we see us performing very well that we want to keep doubling down in. And then there's certainly going to be incrementally more around driving the story of content and AI in the market and again, responding to a lot of the demand and interest that's out there around having solutions that let you tap into your unstructured data. So that all combined will be where we drive the overall go-to-market engine.

Dylan Smith

Yes. And then -- this is Dylan. On the respective economics of direct versus indirect customers, those actually end up being extremely comparable. And that is to say that even when factoring in with whatever margin or payments the partner is getting still nets out to be fairly neutral because in a lot of cases, those customers are getting other benefits, cost leveraged working through that partner, whether it's marketplace credits or part of a broader spend with them or something like that, and tends to come in as such at kind of on average a slightly higher pricing, which kind of offset some of the margin impact and that doesn't even speak to a lot of the other support that our partners provide, whether that's on the customer support side certainly going out and finding these customers, which is more efficient with respect to our own sellers or marketing spend doing the same job.
So I'd say overall, and what's actually a little bit unique, I think, the Box's business versus what you might normally see across the software landscape is our indirect customers are actually effectively just as profitable as our direct customers.

Operator

Lucky Schreiner, DA Davidson.

Lucky Schreiner

Maybe just on the guidance, it seems to imply a little bit of a change in terms of the seasonality of the business being more back-end weighted. Is that the right way to think about it? Or how should we sort of think about linearity here?

Dylan Smith

Yeah. So do you -- for which metric are you asking about specifically?

Lucky Schreiner

Revenue.

Dylan Smith

Revenue? Sure. Yes. So I would say a couple of things there. I mean, first, as we've mentioned, Q1 is kind of artificially lower as it were because of the impact of the leap year and that kind of 120-basis-point impact.
But even adjusting for that, you're right that we do expect to see a gradual acceleration in our revenue growth rate apples-to-apples throughout the year as a lot of -- as we mentioned, a lot of the investments we've been making in growth, kind of the traction for Enterprise Advanced and everything else that Aaron and I have been discussing is what's driving that acceleration and you see that again in the net retention rate and our expected improvement by a point year-over-year as well as with billings coming in a little bit higher than our revenue growth rate expectations for the year.

Lucky Schreiner

And then maybe on the top of funnel interest, have you noticed just in general and just from new customers increasing significantly with the Enterprise Advanced solution? Or is it really more from the existing customer base today given how early it is?

Aaron Levie

Yes. I think given the hourly stage, the plan is, it's heavily weighted to our customers that we're in conversation with and highlighting new use cases. But maybe the plan aside for 1 second, I would say our ability to draw in a wider audience is increasing as a result of the latest functionality that we've introduced. AI has been this incredibly important technology for us in, let's say, expanding the aperture of our potential customer base.
There's a large amount of customers that sort of know that they're sitting on unstructured data, documents and AI is the first time that they are starting to realize they can get insights from that data or automate business processes around that.
So just as an example, I'm often talking to Chief Data Officers and that was never really an audience that we would talk to before, we were usually in kind of core infrastructure within the CIO organization. But the Chief Data Officer realizes now that once you have AI on unstructured data, they can treat that as another data type to pull business insights from. And so this is opening up that aperture as an example.
We're having more conversations with company CTOs that are trying to deliver better experiences and they know that if they can get data from within their unstructured information, they can go and automate a better client-facing experience. And so what's going to happen is as we really, really own this content and AI message in the market, I believe you're going to see a widening of the funnel of the kind of audiences we can go talk to.
Not even to mention in the lines of business that are also trying to figure out how can I bring AI to marketing, how do I bring AI to HR processes, and this is where you're going to see us have a very strong story around how companies can run in an AI-first way really in any line of business that they're operating in.

Operator

Taylor McGinnis, UBS.

Taylor McGinnis

Maybe first just on the operating income margin guide for 2026. So roughly implying flat year over year. I know it sounds like a 60-basis-point headwind embedded in that. But just when we think about that in the context of like where you guys stand with investments, what you're doing around AI, like anything you can give us a little bit more color there in terms of what might be embedded? And then also to just how that relates to comments around significant margin expansion still left in the business.

Dylan Smith

Yeah, sure. I think, Taylor, I would say at a high level, as you mentioned, just given the state of kind of ramping up and making sure that we really both drive the innovation around our product offering as well as invest in things that we started the past year, such as really standing up and setting up our partner ecosystem for success, in addition to kind of investing in some of the key markets, kind of sales segments that we've been seeing really strong results.
We want to make sure that we're doubling down there just given how critical of a juncture it is in terms of a lot of these technology decisions. And with AI, would think about it as really kind of making a lot of those sales and marketing investments. But as you know, even once we have those resources ramped just it takes a few quarters for those deals to come to fruition and then the revenue kind of falls in the year that follows actually closing the deal.
So you can think of that as kind of the lag of a lot of the investments we're making that we are seeing early signs of those paying off, you're not quite getting the revenue impact that we expect to show up over time. And so that's -- a lot of it without doing anything dramatically different is what leads to that confidence and how we're thinking about that approach, which is why this coming year because of that timing, we are expecting to generate again some apples-to-apples operating margin expansion on a year-over-year basis, but nothing like we have generated in the past couple of years, both because the significant improvements that we made on gross margin already, which will incrementally improve over time as well as OpEx leverage being positive, but fairly minimal for this year.

Aaron Levie

Yes. And maybe just to add just one second of color that Dylan covered is philosophically, I think we are -- we're in a moment right now where we're just seeing a tremendous amount of opportunity on really connecting AI and content together and it's the most excited that we've been as a company and from a technology standpoint and the amount of interest from customers is certainly the highest for what they can now do with their data.
So we want to be extremely prudent to be clear, which is why we're making extremely thoughtful investments, and we are still on the path to the higher operating margin levels that we've discussed, but this is a kind of period where we do want to make sure that we're tuning that allocation just a little bit differently than maybe we have in the past couple of years.

Taylor McGinnis

And then just maybe on the outperformance in billings in 4Q. I think you guys had a tougher compare. So that was a really solid number. Could you just maybe talk about what drove that? And then when we look into the first half of the year, the stronger 1Q number versus 2Q? Anything just to flag on that?

Dylan Smith

Yes. So I'd say -- and actually, the response is somewhat related. We'd say in Q4, we had a solid result in the prior quarter, but not an unusually hard compare.
There were some FX headwinds that created a lot of the kind of comparison challenges that we had expected in [DC], FX was actually a little bit more of a factor and a bigger negative impact than we had expected when we set our Q4 expectations, I would say it was really -- that strength of the combination, as mentioned, of just a strong adoption and early momentum with Enterprise Advanced and just overall, a solid bookings quarter for us. And on the early renewal dynamic that I mentioned, that had about a $5 million impact to Q4's results.
So a little less than 2 points of the outperformance was attributable to that. So again, I would say that any way you slice it, we're really pleased with Q4's billing results and came in ahead of our expectations, but did just want to be transparent that there were a couple of items and the payment durations in particular, that kind of helped with the Q4 billings growth rate.
And then turning to Q1 in Q2 is kind of somewhat follows that and would note that in Q1 versus Q2 of last year, that was where we did have both respectively, the weakest and then the strongest billings growth rate at kind of basically flat to 10% plus going Q1 to Q2. At the same time, as we mentioned, there is the impact of the of FX that flips where in Q1 of last year, in Q1 of this coming year, we expect x to be a 300-basis-point tailwind.
But then for Q2, we expect that to be a headwind of about 140 basis points because there were significant FX volatility throughout the first half of last year. So that's one big factor, a nearly 500 basis point swing attributable to FX and then the final piece in Q2 is some of those early renewals that we mentioned in Q4 are actually coming out of customers who had been set to renew in Q2. So that's where we see a little bit of that dynamic set in as well, which is why you see teens growth in Q1 and then comes down a bit in Q2. But that's basically primarily optics and just some of the cosmetic factors.
And again, for the full first half of the year, we expect that billings growth rate to be about 7%, consistent with both the full-year and second-half expectations.

Operator

Pinjalim Bora, JPMorgan.

Pinjalim Bora

Aaron, interesting to see the consumption AI units that you launched. Can you help us understand maybe what drives the consumption of an AI unit? I'm trying to understand if that's like a per document of metadata extraction or that per word of metadata extraction? How granular is it? Can just understand how should we think about kind of the ramping of those AI units?

Aaron Levie

Yeah. So we realize that we probably -- and we'll have somewhere between, I don't know, 5 and 10 different kind of applied AI use cases in making up a time frame for the next year or two, let's just say, and we didn't want to have to have a SKU for each of those different capabilities where a customer has a different billing sheet for each product they want. And then you have a complex sell cycle for every single thing they want to do with AI.
So we looked at the industry, looked at different models and come up with AI units, which is really the substraction layer from any individual sort of function that you want to leverage and so there will be basically a conversion chart that says one AI unit equals x amount of things in that particular capability set. So you could imagine a scenario where extracting data from a page of a document costs a certain number of AI units and then how many ever pages you want to do or how many documents you want to do, that will generate a certain AI consumption amount.
Similarly, you can imagine an a genetic workflow in the future where an agent is sort of doing a long-running process. maybe that takes 5 or 10 minutes for that agent to complete the work, and we would have some kind of conversion ratio on the amount of the amount of kind of work the AI agent is doing and the amount of AI units.
And then one more X factor in there is the model capability or quality that you want to be able to leverage becomes yet another variable. Also a lower class model that might be great at one kind of simple thing would cost a certain number of AI units, but then a much more powerful reasoning model that's doing thinking and chain of thought in the response that might cost a little bit more. So those are the different kind of philosophical variables.
I'd say we're way too early to ask anybody to start to model that. This was just released a couple of weeks ago as the first way that we can help our customers have more flexibility in how they're using AI and then we're going to expand it from there. But essentially, you can think about it as a customer will call us a -- I have 100,000 documents I want to process, and then we'll convert that into a number of AI units that they need in addition to the core functionality, which is unlocked at the Enterprise Advanced plan in this case.

Pinjalim Bora

And then, Dylan, maybe help us understand about the federal sector. How should investors kind of think about the US federal exposure for Box? And are you baking in some caution around that segment as you're thinking about the guidance?

Dylan Smith

Sure. So some of the dynamics with that market are what leads to the prudence. And so we are certainly want to be cautious. But again, I think it's too soon to tell exactly what the impact could look like, and that has been such a key area for our growth where we've seen a lot of green shoots because of becoming FedRAMP high compliance and all of those things that we are certainly remain focused on that opportunity.
What I would say in terms of the size and impact in terms of our current business, fairly minimal exposure there where what I would say is that our total United States public sector business is in the mid-single-digit percentage range. Our overall revenue and then the federal business is a portion of that as we also have a pretty successful state and local business as well. So it's kind of low single-digit percentage range of overall revenue, but certainly one that we are paying a lot of attention to.

Operator

Rishi Jaluria, RBC Capital Markets.

Rishi Jaluria

Wonderful. Maybe I want to start one a little bit housekeeping and better understanding for Dylan and then one for Aaron. Dyl, just to start off 1with, you saw a pretty big increase in long-term RPO and you called out duration as kind of being a tailwind. At the same time, we see a lot of Enterprise software companies that are actually seeing shorter and a shortening duration. Can you maybe walk us through the dynamic of why you're seeing better contract duration?
And then I have a follow-up for Aaron.

Dylan Smith

Yes. I would say really maybe a function of where we're coming from and that as customers are increasingly viewing especially injunction an enterprise previously plus where we had started to see this trend. And now with Enterprise Advanced, that really represents the type of commitment and kind of demonstration buy into the platform that it makes sense to sign a multiyear deal, whereas maybe those customers had just had annual contracts as well, particularly as some of these use cases that are more complex and add more value takes some time to ramp up.
And so built into that contract structure, there might be a gradual increase in the amount that they're getting billed per year. So there's contractually some incentive and really alignment for customers to adopt our solutions as quickly as possible, that is having an impact there.
I would say the other thing that fuels it, if not just the overall duration, but also the volume of early renewals, right? So if you had a customer where maybe they had a year remaining even on a multiyear contract, and now they're renewing early in a given period, for example, because they're really excited and want to get all the capabilities of Enterprise Advanced as soon as possible now instead of having -- with a year to last, it would have had nothing in long-term RPO.
And now if they re-up for another three-year term, early renewing that has now two years that shows up in backlog and our long-term RPO growth. So I'd really say it's the combination of just becoming a much more critical and strategic perform for our customers, which leads to that lengthening as well as some of the mechanics of the early renewals, both fueling that growth.

Rishi Jaluria

And then, Dylan, if we really just think about -- you mentioned during your prepared remarks that you are using Box internally as well as some of the Box AI features and SKUs. Can you maybe talk a little bit about what have you seen internally, especially anything that can be benchmarked in terms of drive greater efficiency, cost savings or even kind of driving better execution and better growth as a result of using Box AI, what would have you seen anything you'd benchmark? Thanks.

Aaron Levie

Yeah. So this is Aaron. But we have a very kind of box-on-box first approach to AI rollout. So we want Boxers to be in the product for every possible use case so we can figure out ahead of our customers what's working well, what's not. So we're getting a stream of great anecdote and quantitative case studies.
And so we still have to aggregate kind of all of these across the board, but we'll have examples of just a single HR process where we're using Box hubs and rolling that out for employees to ask questions about, let's say, a new kind of HR process, just saving dozens and dozens of hours for one micro type of tasks like that, that normally would have been internal e-mails and conversations constantly from people asking questions.
Now you offload on to AI. So we have hundreds of those scenarios. And we'll be sharing a bit more at the Analyst Day to kind of roll that up. But just think about it as whether you're interacting with documents or contracts or writing a product spec on a team or working on a marketing asset or asking questions about HR information or sales materials or product support questions, all of that is routing through Box AI today.
And so the expectation, of course, is that we're going to be shipping more software. We're going to be selling to customers better. We're going to be able to deliver more marketing campaigns, and that is that's already sort of showing up across the business.

Operator

Josh Baer, Morgan Stanley.

Josh Baer

Great. We're seeing a really strong uplift in Advanced from Enterprise Plus. So existing customers clearly finding the value and all the new features in Box AI. I was hoping you could talk about how Box AI, the latest innovation are impacting new customer wins. And also just wondering if you've seen all this new technology as an accelerant to broader platform enterprise content management platform rip and replace as well.

Aaron Levie

Yes. So maybe to start with the second question first. AI is absolutely going to be a catalyst for driving more more, let's say, re-platforming of traditional approaches to document management and content management. And so we sort of see two kind of vectors to this. The first is a customer that might be on a pure legacy enterprise content management system and just the amount of -- the modernization of the functionality is just not there to be able to go drive intelligence from data.
We have a partner that we're building more and more momentum with that that has sort of highlighted how little innovation there's been in this category from maybe traditional players and it's creating a huge opening for us because now Box AI is letting them start to bring us in conversations that traditionally, they've not been able to move to the cloud or leverage a modern solution for, but now AI kind of creates that catalyst for them. So that's sort of Vector 1.
And then Vector 2 is think about all of the business units or parts of the organization that didn't have a traditional content management system. And the customer wasn't really even thinking they were in the market for traditional enterprise content management. So the team marketing -- leveraging or managing digital assets or the team that was managing contracts or the invoice processing team in an organization, a lot of times, that data just goes into file systems or it's stuck in e-mail attachments.
And so AI now all of a sudden dramatically exceeds what they were doing in their workflows before, which has the opportunity now to cause customers to say, okay, maybe actually, I do want software for this problem I don't want to just have my data floating around and handling these workflows on a manual basis. So that's going to be frankly, I think over the medium term, the largest part of the category, the traditional enterprise content management space is measured in the sort of $7 billion to $10 billion range, plus or minus, depending on the analyst you look at, but the much bigger opportunity is actually all of the use cases that today don't go into an OpenText or a documentum or a Highland and that's the data that we can now go drive a significant amount of automation for.
And then to the first question, we are, again, very, very early in rolling out Enterprise Advanced. A lot of the demand is coming from existing customers, but we're now going to make sure that this is very much front and center for all of our new sales conversations, all of our marketing campaigns. Our entire marketing story right now is content plus AI and really highlighting what customers can now do with it and enterprise Advanced is certainly going to be the best plan type to enter for driving those use cases.

Operator

That concludes our Q&A session. I will now turn the conference back over to Cynthia Hiponia, VP of Investor Relations, for closing remarks.

Cynthia Hiponia

Great. Thank you, everyone. As a reminder, we're holding our fiscal '26 Financial Analyst Day on Tuesday, March 18, just reach out to Elaine or myself at ir@box.com to register. And we look forward to chatting with you again on Analyst Day and on our next earnings call.

Operator

That concludes today's conference call. You may now disconnect.